Power generation control method and apparatus based on active and reactive power regulation of photovoltaic power generation

Abstract

Disclosed in the present invention are a power generation control method and apparatus based on active and reactive power regulation of photovoltaic power generation. The method comprises: acquiring a planned power generation curve allocated to a photovoltaic power plant by a power grid side and a predicted power generation curve generated by the photovoltaic power plant; determining whether the predicted power generation curve is fitted with the planned power generation curve; if the predicted power generation curve is not fitted with the planned power generation curve, modifying the planned power generation curve by means of deep learning and re-sending same to the photovoltaic power plant; if the predicted power generation curve is fitted with the planned power generation curve, determining whether a measured power generation curve of active power and reactive power of the photovoltaic power plant is fitted with the predicted power generation curve; if the measured power generation curve is not fitted with the predicted power generation curve, modifying a panel cleaning scheme of the photovoltaic power plant; and if the measured power generation curve is fitted with the predicted power generation curve, the photovoltaic power plant performing power generation control on the basis of safety requirements for power plant equipment and the planned power generation curve. By means of two successive fitting determinations, the planned power generation curve approaches the predicted power generation curve, and the measured power generation curve approaches the predicted power generation curve, thereby improving the certainty of photovoltaic power generation control, ensuring the service life and working performance of equipment, and overcoming the difficulty in power generation control of the photovoltaic power plant.

Description

A power generation control method and device based on active and reactive power adjustment of photovoltaic power generation

[0001] This invention claims priority to Chinese Patent Application No. 202411896392.4, filed on December 20, 2024, with the application title “A Power Generation Control Method and Device Based on Active and Reactive Power Adjustment of Photovoltaic Power Generation”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This invention relates to the field of power system technology, and more specifically to a power generation control method and device based on the adjustment of active and reactive power of photovoltaic power generation. Background Technology

[0003] In power systems, grid control controls frequency through active power and voltage through reactive power. Power plants often need to coordinate their power generation with the grid's demand. This often involves requiring power plants to generate more active or reactive power. Due to the high uncertainty in power generation from certain types of power plants (e.g., photovoltaic power plants), and the differences in the characteristics of equipment used in different types of power plants, the difficulty of power generation control in power systems is further increased. Improper control can easily lead to damage to power plant equipment, insufficient power generation to meet grid demand, or reduced controllability of power generation or electricity consumption costs. Currently, how to achieve effective power generation control in photovoltaic power generation scenarios with high uncertainty is a pressing technical problem that needs to be solved in power systems. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a power generation control method and apparatus based on the adjustment of active and reactive power in photovoltaic power generation. The aim is to provide a power generation control scheme for photovoltaic power plants, effectively overcoming the difficulties in power generation control.

[0005] The present invention discloses the following technical solutions:

[0006] The first aspect of this invention provides a power generation control method based on active and reactive power adjustment of photovoltaic power generation, the method comprising:

[0007] Obtain the planned power generation curves of active and reactive power allocated to the photovoltaic power plant by the grid, and obtain the predicted power generation curves of active and reactive power generated by the photovoltaic power plant itself.

[0008] Determine whether the predicted power generation curve fits the planned power generation curve;

[0009] If the predicted power generation curve cannot be fitted to the planned power generation curve, the grid side modifies the planned power generation curve through deep learning and resends it to the photovoltaic power plant after the modification is completed; if the predicted power generation curve fits the planned power generation curve, it is determined whether the measured power generation curves of the photovoltaic power plant's active power and reactive power fit the predicted power generation curve.

[0010] If the measured power generation curve cannot be matched with the predicted power generation curve, the panel cleaning scheme of the photovoltaic power plant is modified; if the measured power generation curve matches the predicted power generation curve, the photovoltaic power plant controls power generation based on the safety requirements of the power plant equipment and the received planned power generation curve.

[0011] In an optional implementation, the photovoltaic power plant controls power generation based on the safety requirements of the power plant equipment and the received planned power generation curve, including:

[0012] Determine whether the photovoltaic power plant meets the safety requirements of its equipment if it performs active and reactive power conversion control during the power generation process according to the planned power generation curve.

[0013] If the safety requirements are met, the photovoltaic power plant controls power generation based on the planned power generation curve; if the safety requirements are not met, the photovoltaic power plant controls power generation based on the planned power generation curve using peak shaving technology.

[0014] In an optional implementation, determining whether the safety requirements of the photovoltaic power plant's equipment are met if the photovoltaic power plant performs active and reactive power conversion control during power generation according to the planned power generation curve includes:

[0015] If the photovoltaic power plant performs active and reactive power conversion control during the power generation process according to the planned power generation curve, determine whether the current and temperature of the photovoltaic power plant's equipment meet safety requirements, and whether the operating voltage of the photovoltaic power plant's equipment is normal.

[0016] In an optional implementation, the photovoltaic power plant controls power generation based on the planned power generation curve using peak shaving technology, including:

[0017] Adjust the conduction angle of the thyristor in the photovoltaic power plant to perform peak reduction.

[0018] In an optional implementation, the method further includes the following steps before power generation control:

[0019] The characteristics of active and reactive power variation of the inverter of the photovoltaic power plant under the influence of voltage factors are analyzed, and the inflection point information in the curves of active and reactive power variation characteristics under different voltages is determined. The inflection point information includes the values ​​on the reactive power axis and the active power axis.

[0020] Power generation control includes:

[0021] The reactive power value of the inverter is controlled to be less than the reactive power coordinate axis value in the inflection point information.

[0022] In an optional implementation, the grid side modifies the planned power generation curve using deep learning, including:

[0023] The grid-side deep learning method learns the difference between the measured daily power generation of the photovoltaic power plant and the historical daily power generation, and learns the difference between the comprehensive efficiency of the photovoltaic power plant and the comprehensive efficiency of the same period in history. It also combines the panel cleaning degree data of the photovoltaic power plant to identify the impact of haze and adjust the planned power generation curve.

[0024] In an optional implementation, the modification of the photovoltaic power plant's panel cleaning scheme includes:

[0025] Analyze the difference between the measured daily power generation and the historical daily power generation of the photovoltaic power plant;

[0026] Analyze the correlation between the panel cleaning level data and power generation efficiency in each zone of the photovoltaic power plant;

[0027] Based on the differences in power generation, the corresponding relationship obtained by analysis is obtained, the difference between the comprehensive efficiency of each zone of the photovoltaic power plant and the comprehensive efficiency of each zone in the same period of history, and the difference between the overall comprehensive efficiency of the photovoltaic power plant and the overall comprehensive efficiency in the same period of history, so as to obtain the degree of impact of haze on each zone.

[0028] The current board cleaning plan has been modified based on the degree of impact of fog on each zone.

[0029] In an optional implementation, modifying the current board cleaning scheme based on the degree of impact of fog on each partition includes:

[0030] Based on the degree of impact of fog on each zone, panel cleaning recommendations are generated; the panel cleaning recommendations include at least one of the following: area cleaning recommendations, cleaning frequency recommendations, cleaning speed recommendations, and cleaning manpower allocation recommendations;

[0031] Modify the current board cleaning plan based on the board cleaning suggestion information.

[0032] In an optional implementation, the planned power generation curves for active and reactive power allocated to the photovoltaic power plant by the grid side are generated in the following way:

[0033] Based on the characteristic production factor data of each power plant and the control production factor data of the power grid, the grid side determines the load curve to be allocated to each power plant, including the planned power generation curves of active power and reactive power allocated to photovoltaic power plants. The active power control frequency is used, and the reactive power control voltage is used.

[0034] In optional implementations, the photovoltaic power plant generates predicted power generation curves for active and reactive power in the following ways:

[0035] The photovoltaic power plant generates predicted power generation curves for active and reactive power based on equipment performance, weather data, and historical power generation for the same period. The equipment performance includes the characteristics of the inverter's active and reactive power changes under the influence of voltage factors.

[0036] In an optional implementation, before the photovoltaic power plant performs power generation control based on the safety requirements of the power plant equipment and the received planned power generation curve, the method further includes:

[0037] The photovoltaic power plant sends power and cost data for the curtailment assistance function to the grid side.

[0038] If the photovoltaic power plant receives a control command from the grid side, it will control power generation based on the safety requirements of the power plant equipment and the received planned power generation curve; the control command is sent by the grid side after reaching a transaction agreement with the photovoltaic power plant based on the power data and the cost data.

[0039] A second aspect of the present invention provides a power generation control device based on active and reactive power adjustment of photovoltaic power generation, the device comprising:

[0040] The curve acquisition module is used to acquire the planned power generation curves of active and reactive power allocated to the photovoltaic power plant by the grid side, and to acquire the predicted power generation curves of active and reactive power generated by the photovoltaic power plant itself.

[0041] The fitting judgment module is used to determine whether the predicted power generation curve fits the planned power generation curve.

[0042] The curve modification module is used to modify the planned power generation curve through deep learning if the predicted power generation curve cannot be fitted to the planned power generation curve, and then resend the modified curve to the photovoltaic power plant.

[0043] The fitting judgment module is further used to determine whether the measured active power and reactive power curves of the photovoltaic power plant fit the predicted power generation curve if the predicted power generation curve fits the planned power generation curve.

[0044] The cleaning scheme modification module is used to modify the panel cleaning scheme of the photovoltaic power plant if the measured power generation curve and the predicted power generation curve cannot be fitted.

[0045] The power generation control module is used to control the power generation of the photovoltaic power plant based on the safety requirements of the power plant equipment and the received planned power generation curve if the measured power generation curve fits the predicted power generation curve.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] This invention discloses a power generation control method and apparatus based on active and reactive power adjustment of photovoltaic power generation. The method includes: acquiring the planned power generation curves of active and reactive power allocated to the photovoltaic power plant by the grid side, and acquiring the predicted power generation curves of active and reactive power generated by the photovoltaic power plant itself; determining whether the predicted power generation curves fit the planned power generation curves; if they cannot fit, deep learning is used to modify the planned power generation curves and resend them to the photovoltaic power plant; if they fit, it is determined whether the measured power generation curves of the photovoltaic power plant fit the predicted power generation curves. If the measured power generation curves cannot fit the predicted power generation curves, the panel cleaning scheme of the photovoltaic power plant is modified; if the measured power generation curves fit the predicted power generation curves, the photovoltaic power plant performs power generation control based on the safety requirements of the power plant equipment and the received planned power generation curves. In this invention, by analyzing whether the predicted power generation curves fit the planned power generation curves, and in cases where they cannot fit, allowing the grid side to continue deep learning to adjust the planned power generation curves and resend them to the photovoltaic power plant, the load plan of the grid side can more stably tend towards the power generation trend predicted by the photovoltaic power plant itself. By analyzing whether the measured power generation curve fits the predicted power generation curve, and adjusting the panel cleaning scheme of the photovoltaic power plant accordingly, the measured results can be improved, allowing subsequent measured power generation curves to closely approximate the predicted curves, thus greatly enhancing the determinism of photovoltaic power generation control in the power system. The photovoltaic power plant's power generation control is based on the safety requirements of the power plant equipment, ensuring equipment lifespan and performance. Therefore, this invention addresses the needs of both the power plant and the grid, effectively overcoming the challenges of photovoltaic power generation control. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 is a schematic diagram of a power system scenario provided by an embodiment of the present invention;

[0050] Figure 2 is a flowchart of a power generation control method based on active and reactive power adjustment of photovoltaic power generation provided by an embodiment of the present invention;

[0051] Figure 3 is a schematic diagram of the characteristic curves of an inverter;

[0052] Figure 4 is a curve comparison chart provided by an embodiment of the present invention;

[0053] Figure 5 is a schematic diagram showing the change of the overall efficiency of a photovoltaic power plant with radiation according to an embodiment of the present invention;

[0054] Figure 6 is an example flowchart of the modified panel cleaning scheme for photovoltaic power plants;

[0055] Figure 7 is a flowchart of another power generation control method based on active and reactive power adjustment of photovoltaic power generation provided by an embodiment of the present invention;

[0056] Figure 8 shows the signaling diagram for the overall power control of the power system;

[0057] Figure 9A is a schematic diagram of segmented control power factor setting;

[0058] Figure 9B is a schematic diagram showing the comparison of peak reduction effects;

[0059] Figure 10 is a schematic diagram of a power generation control device based on the adjustment of active and reactive power of photovoltaic power generation according to an embodiment of the present invention. Detailed Implementation

[0060] Figure 1 is a schematic diagram of a power system scenario. As shown in Figure 1, power plants of various power generation types (such as photovoltaic, thermal power, wind power, nuclear power, hydropower, etc.) may communicate with the grid side and transmit power. These different types of power sources can be collectively referred to as the power plant side. In this embodiment of the invention, the power plant side can report its own data to the grid side and provide timely feedback; the grid side can perform scheduling on the power plant side, such as scheduling of voltage, frequency, power, etc. In the actual power production process, the grid side and the power plant side need to perform close scheduling to achieve better and more stable power supply and maintain the cost requirements of both parties. In actual scenarios, the configuration of the power plant side is not limited to the types of power plants shown in Figure 1. Depending on the differences in the level of power development, basic power infrastructure, and abundance of natural resources in various countries, the power supply structure of some countries' power systems may only include some of the types of power plants shown in Figure 1.

[0061] As mentioned above, the power generation of photovoltaic (PV) power plants is highly uncertain due to factors such as weather and seasons, making power generation control difficult. Currently, there is a lack of effective power generation control solutions. To address this, the inventors propose the following technical solution: First, obtain the planned power generation curve allocated to the PV power plant by the grid and the predicted power generation curve generated by the PV power plant; determine whether the predicted power generation curve fits the planned power generation curve (first fitting judgment). If they cannot fit, deep learning is used to modify the planned power generation curve and resend it to the PV power plant; for a second fitting judgment, determine whether the measured active and reactive power curves of the PV power plant fit the predicted power generation curve. If they cannot fit, modify the PV power plant's panel cleaning scheme; for a third fitting judgment, the PV power plant controls power generation based on the plant's equipment safety requirements and the planned power generation curve. In the first fitting judgment, if they do not fit, the grid side makes changes: modifying the planned power generation curve through deep learning to make it approximate the predicted power generation curve, promoting consensus between the two sides on subsequent PV power generation. In the second fitting judgment, by changing the PV panel cleaning scheme, the measured power generation curve approximates the predicted power generation curve. The above measures enhance the certainty of photovoltaic power generation control, ensure equipment lifespan and performance, overcome the difficulties in photovoltaic power plant power generation control, and effectively solve the current technical problems.

[0062] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0063] Figure 2 is a flowchart of a power generation control method based on active and reactive power adjustment of photovoltaic power generation according to an embodiment of the present invention. As shown in Figure 2, the power generation control method based on active and reactive power adjustment of photovoltaic power generation provided by the present invention includes:

[0064] S201. Obtain the planned power generation curves of active and reactive power allocated to the photovoltaic power plant by the grid side, and obtain the predicted power generation curves of active and reactive power generated by the photovoltaic power plant itself.

[0065] In this embodiment of the invention, the grid side can determine the load curve to be allocated to each power plant based on the characteristic production factor data of each power plant and the control production factor data of the grid side. This includes the planned power generation curves of active power and reactive power allocated to the photovoltaic power plant, wherein the active power control frequency is used and the reactive power control voltage is used.

[0066] The basis for classifying production factors on the power plant side is called the characteristic production factor group; the basis for classifying production factors on the power grid side is called the control production factor group.

[0067] The characteristic production factor group includes multiple levels of characteristic production factors. Specifically, these characteristic production factors can be subdivided into: primary characteristic production factors, secondary characteristic production factors, and tertiary characteristic production factors. Primary characteristic production factors are natural characteristic production factors (or natural indicators), which directly reflect natural characteristics. Secondary characteristic production factors are single-system characteristic production factors (or single-system indicators), which are production factors related to individual systems within the power system. Tertiary characteristic production factors are plant system characteristic production factors (or plant system indicators), which are production factors related to the overall system within the power system. As shown in Figure 1, power plants vary in type, such as thermal power plants, photovoltaic power plants, wind power plants, and hydropower plants, each employing different power generation methods, using different natural resources, and having different internal system structures. To achieve more accurate digital power analysis, this invention's technical solution can subdivide primary, secondary, and tertiary characteristic production factors for various power generation types. Therefore, for each type of power plant, data for the corresponding level of characteristic production factors can be extracted subsequently.

[0068] This invention primarily focuses on photovoltaic (PV) power plants. The characteristic production factor data of PV power plants used by the grid mainly involves primary and tertiary characteristic production factors. Taking PV power generation as an example, primary characteristic production factors may include PV irradiance, and tertiary characteristic production factors may include PV efficiency.

[0069] The control production factor group includes multiple aspects of control production factors. Control production factors are production factors that reflect power grid performance and are composed of some key physical quantities in the power system. Control production factors help provide an effective theoretical basis for power system control. In this embodiment of the invention, the control production factors involved may include the relationship between active power, reactive power, and voltage. The data form of control production factors can be (voltage, active power, reactive power). In a power system, voltage and frequency are not directly related, but they are related to power (active power and reactive power, respectively), creating an indirect relationship. When reactive power is insufficient, the voltage drops; when reactive power is excessive, the voltage rises. When active power is insufficient, the frequency drops; when active power is excessive, the frequency rises.

[0070] The characteristic production factor data of power plants can be transmitted to the grid side, allowing the grid side to promptly understand the specific values ​​of these data. Furthermore, the data can be processed digitally into vector form for deep learning and application. For photovoltaic (PV) power plants, the grid side can determine the load curve to be allocated to the PV power plant—that is, the planned generation curve of active and reactive power—based on the PV power plant's characteristic production factor data (PV irradiance, PV efficiency, etc.) and the grid side's control production factor data.

[0071] As an example, the grid side establishes an illuminance data vector based on the 2023 photovoltaic illuminance data provided by the power plant side. Based on this vector and control production factor data, a mapping function is constructed, and this mapping function is used to predict grid power generation. Furthermore, based on data-driven incentives and the data provided by the power plant side, the active and reactive power trends of the grid are predicted. Based on the mechanisms of the grid and power generation equipment, a power generation control logic is formed on the grid side. For example, reactive power at the power plant side is increased when voltage decreases, and active power at the power plant side is increased when frequency decreases. In addition, there may be a certain correlation between the capacity and frequency of different types of power generation equipment; the grid side can also incorporate this correlation to comprehensively consider the planned power generation curves allocated to each power plant.

[0072] Furthermore, the grid side can also generate planned power generation curves by combining equipment mechanisms (or equipment characteristics). Taking a photovoltaic power plant as an example, in the active and reactive power variation characteristic curves (PQ curves) of the inverter, the horizontal axis represents reactive power and the vertical axis represents active power. Figure 3 is a schematic diagram of an inverter characteristic curve. As shown in Figure 3, different curves represent the trend of active power changing with reactive power of the same inverter at different voltages. In the figure, Un represents voltage, PU is a per-unit value, dimensionless, Q represents reactive power in kVar, and P represents active power in kW. It can be seen from the figure that in the first quadrant, as the horizontal axis increases, the vertical axis tends to gradually decrease. And after reaching a certain position, the trend of the vertical axis decreasing with the increase of the horizontal axis is too obvious (which can be understood as the slope being too large, or the slope exceeding a certain threshold). This significant turning point on the curve is called the inflection point. The grid side can determine the inflection point information that the increase in reactive power leads to an excessive decrease in active power, and give corresponding planned power generation curves for different power plant equipment mechanisms. During this process, the effects of voltage and frequency can be taken into account, as well as the characteristics of the equipment, so that the generated planned power generation curve controls reactive power without excessively affecting active power.

[0073] The power plant also combines equipment performance and historical power generation data to generate predicted power generation curves. Optionally, for photovoltaic (PV) power plants, the plant can generate predicted active and reactive power curves based on equipment performance, weather data, and historical power generation data. The equipment performance data also includes the active and reactive power variation characteristics of the inverters used in the PV power plant under voltage factors, as shown in Figure 3.

[0074] For ease of understanding, example forms of planned and predicted power generation curves are shown in the accompanying drawings. Figure 4 is a curve comparison diagram provided by an embodiment of the present invention. In the figure, the horizontal axis represents time, and the vertical axis represents power generation (unit: kW·h). The curve with diamond-shaped nodes represents the planned power generation curve generated by the power grid, the curve with X-shaped nodes represents the predicted power generation curve generated by the photovoltaic power plant, and the curve with circular nodes represents the measured power generation curve of the photovoltaic power plant. A point on the curve represents the power generation on a specific day of a specific month.

[0075] Understandably, the planned power generation curve represents the grid's planned expectations, while the predicted power generation curve represents the power plant's own achievable expectations. To better control the power generation of photovoltaic power plants, a common understanding between the grid and the power plant is necessary. Conversely, if the planned and predicted power generation curves deviate significantly, on the one hand, the power plant may be unable to respond to grid dispatch, and the two sides may be unable to reach a consensus on electricity prices and quantities to facilitate transactions; on the other hand, even if the power plant responds to the grid's requirements, they may ultimately fail to meet them. In some cases, this could even damage the power plant's equipment, significantly reducing its lifespan and increasing power generation costs. Therefore, a fitting analysis can be used to determine whether the two curves fit, and then appropriate measures can be taken based on the assessment. This will be discussed in conjunction with sections S202 to S204 below.

[0076] S202. Determine whether the predicted power generation curve fits the planned power generation curve. If the predicted power generation curve and the planned power generation curve cannot fit, proceed to S203; if the predicted power generation curve and the planned power generation curve fit, proceed to S204.

[0077] In this embodiment of the invention, the fitting judgment between the predicted power generation curve and the planned power generation curve can be referred to as the first fitting judgment in the technical solution process of this embodiment. The first fitting judgment mainly analyzes whether the difference between the predicted power generation curve and the planned power generation curve is too large. If they cannot be fitted, it means that the curve difference is too large; if they can be fitted, it means that the curve difference is within an acceptable range.

[0078] Besides visual analysis to determine whether two curves fit, other automated fitting evaluation methods can be used in this step. For example, statistical methods such as residual analysis, R-squared, least squares, root mean square error, and mean absolute error can be used to evaluate the fit between the two curves. No restrictions are placed on the statistical tools or comparison methods used for this judgment.

[0079] If the predicted power generation curve fits the planned power generation curve, it means that the planned power generation curve on the grid side is feasible. Then, the subsequent operation process can be executed, proceeding to S204, where the power plant side performs a second fitting judgment. If the predicted power generation curve does not fit the planned power generation curve, it means that the planned power generation curve on the grid side lacks feasibility. In this case, S203 can be implemented, allowing the grid side to proactively modify the planned power generation curve, improving the feasibility of the modified planned power generation curve. The following section introduces and explains the two processing branches after the first fitting judgment, in conjunction with S203 and S204.

[0080] S203. The grid side modifies the planned power generation curve through deep learning and resends it to the photovoltaic power plant after the modification is completed.

[0081] In this specific implementation, the power grid side can deeply learn the difference between the measured daily power generation of the photovoltaic power plant and the historical daily power generation, and learn the difference between the comprehensive efficiency of the photovoltaic power plant and the comprehensive efficiency of the same period in history. Combined with the panel cleaning degree data of the photovoltaic power plant, the impact of fog is identified, and the planned power generation curve is adjusted.

[0082] The measured daily power generation of a photovoltaic (PV) power plant can be obtained through actual measurement and data collection. Historical daily power generation can be obtained by collecting historical data. For example, the historical daily power generation data of the PV power plant in 2022 and 2023 can be collected into a historical data database. In response to the deep learning needs of the grid side, this historical daily power generation data can be retrieved from the database and sent to the grid side. Alternatively, the grid side can also have the authority to query the historical daily power generation data of the PV power plant from the database. When the grid side needs to perform deep learning, it can directly query and retrieve data from the database to achieve its deep learning objectives. Similarly, in addition to the relevant data on daily power generation, the overall efficiency of the PV power plant can also be compared with that of the same period. Figure 5 is a schematic diagram of the change of the overall efficiency of a PV power plant with radiation provided by an embodiment of the present invention. In Figure 5, the horizontal axis represents the radiation, with the unit being kW·h / m². 2 The vertical axis represents overall efficiency, which is dimensionless. Data points of the same shape represent the same month. The dashed lines in the figure are regression lines formed by aggregating data points of the same shape. As can be seen from the dashed lines in the figure, the regression lines fitted to different months are different, which indicates that the trend of overall efficiency changing with radiation may be different in different months.

[0083] In this embodiment of the invention, data points that deviate significantly from the regression line can be removed. Removing some data points ensures the overall reliability of the simulation data, while comparing historical efficiency, current efficiency, and efficiency trends provides excellent guidance and correction for power prediction and intraday trading.

[0084] Similar to Figure 5, the grid side can compare the overall efficiency of photovoltaic power plants this year with the overall efficiency of the same period in history, and then learn about the differences in overall efficiency.

[0085] In this embodiment of the invention, the degree of panel cleaning is also considered during the grid-side deep learning process to modify the planned power generation curve. Referring to Figure 4, the planned power generation curve deviates significantly from the predicted power generation curve at certain times. Furthermore, the measured daily power generation on certain dates differs drastically from the historical daily power generation for the same period, or the overall efficiency of the photovoltaic power plant on certain dates differs drastically from the historical photovoltaic efficiency for the same period. This significant difference in data for the same period may be due to the impact of haze. Generally, haze lasts for a long time and has a significant impact on photovoltaic power generation. Moreover, haze particles cover the surface of the panels, potentially causing a lasting impact on the photovoltaic panels even after the haze ends. Therefore, the grid side can identify the impact of haze by combining the panel cleaning degree data of the photovoltaic power plant. For ease of understanding, the impact of panel cleaning on power generation efficiency is illustrated below with an example.

[0086] The photovoltaic power plant is equipped with transformer substation units T01 and T02. Table 1 shows the radiation levels, power generation of T01, cumulative cleaning ratio of T01, system efficiency of T01 unit, power generation of T02, cumulative cleaning ratio of T02 unit, system efficiency of T02 unit, system efficiency difference between T01 and T02, and power generation efficiency of the two inverters (T01-I26 and T01-I27) of T01 for multiple dates.

[0087] In this example, statistics were collected from December 1, 2020, to January 4, 2021. From December 1, 2020, the system efficiency gradually decreased from 80% to approximately 75%. On December 6, 2020, due to an unintentional disconnection of the network cable by an internet company, some radiation data was lost, distorting the system efficiency. From December 9 to December 15, 2020, there was heavy fog, and garbage and waste were burned at night around the project site, making the system efficiency inaccurate during this period. As shown in Table 1, cleaning of the T01 transformer substation unit began on December 22, 2020, and the cumulative cleaning percentage for T01 began to appear. From December 22, 2020, when component cleaning began, to January 4, 2021, the system efficiency of the T01 unit gradually increased from 76% to 89%, an increase of approximately 13%. T02 was not cleaned during the statistical period, so its cumulative cleaning percentage remained at 0.00%. Before the module cleaning in the T01 transformer substation area (before December 22nd), the power generation efficiency of T01 and T02 was comparable, with a weighted average efficiency difference of only about 1.6%. After module cleaning, the power generation efficiency difference between T01 and T02 transformer substations gradually increased from 1.16% to 11.81%. The power generation efficiency of inverters T01-I26 (cleaned 11 days ago) and T01-I27 (cleaned 12 days ago) did not fluctuate significantly after cleaning. However, it can still be observed that inverter T01-I26 has a slight advantage in power generation efficiency because it was cleaned one day later than inverter T01-I27.

[0088] By comparing the data in Table 1, the following conclusions can be drawn:

[0089] A comparison of the power generation efficiency of units T01 and T02 shows that cleaning the photovoltaic modules significantly improves the system's power generation efficiency, increasing it by approximately 10%. Data from the T01-I26 and T01-I27 inverters over ten days after cleaning indicates no significant decrease in efficiency; the small amount of dust after cleaning has little impact on efficiency in the short term. Following the smoggy weather from December 9th to 15th, 2020, the system efficiency fluctuated considerably. This was primarily due to the discrepancy between the measured radiation data from the meteorological station and the actual radiation received by the photovoltaic panels caused by the smog. Furthermore, after the weather cleared on December 16th, 2020, the smog adhering to the photovoltaic module panels solidified due to sun exposure, which may have further contributed to the decrease in system efficiency.

[0090] The examples above clearly demonstrate that analyzing the cleaning level data of photovoltaic power plant panels can better and more accurately identify the impact of haze. The grid side can also use deep learning to drive adjustments to planned power generation curves based on multi-source data.

[0091] Table 1

[0092] S204. Determine whether the measured power generation curves of the photovoltaic power plant and the predicted power generation curve fit each other. If the measured power generation curves and the predicted power generation curves cannot fit each other, proceed to S205; if the measured power generation curves and the predicted power generation curves fit each other, proceed to S206.

[0093] Based on the introduction in S202 above, if the result of the first fitting judgment is yes, it means that the planned power generation curve on the grid side is feasible. However, it is still necessary to analyze whether the actual measured situation of the photovoltaic power plant matches the expectation. Therefore, a second fitting judgment is performed in S204. Specifically, this judgment determines whether the measured active and reactive power generation curves of the photovoltaic power plant fit the predicted power generation curve. The measured and predicted power generation curves can be understood with reference to Figure 4. In the example in Figure 4, the curve with X-shaped points as nodes represents the predicted power generation curve of the photovoltaic power plant, and the curve with circular points as nodes represents the measured power generation curve of the photovoltaic power plant. The fitting effect of the two curves can be evaluated using statistical methods such as residual analysis, R-squared value, least squares method, root mean square error, and mean absolute error. In possible implementations, the analysis method used in the first fitting judgment is consistent with the analysis method used in the second fitting judgment.

[0094] If the measured power generation curve fits the predicted power generation curve, it means that the deviation between the actual power generation and the expected power generation of the photovoltaic power plant is within an acceptable range. This allows the process to proceed to S206, where the power generation procedure can be implemented in accordance with safety requirements.

[0095] However, if the measured power generation curve cannot fit the predicted power generation curve, it means that the actual power generation of the photovoltaic power plant deviates significantly from the expected power generation. In this case, if power generation continues according to the received planned power generation curve, it will obviously not achieve the planned power generation effect on the grid side. In order to promote the successful and stable completion of power plant transactions, it is necessary to implement corresponding solutions on the power plant side to improve the actual power generation of the photovoltaic power plant. Therefore, we proceed to S205, addressing the issue from the perspective of panel cleaning within the photovoltaic power plant itself. The following section, in conjunction with S205 and S206, introduces and explains the two processing branches after the first fitting judgment.

[0096] S205. Modify the panel cleaning plan of the photovoltaic power plant.

[0097] The modification of the panel cleaning scheme is primarily due to the consideration that the current scheme may not meet the requirements for high-efficiency power generation in terms of speed, frequency, or coverage, leading to a significant deviation between the measured and predicted power generation curves of the photovoltaic power plant. Therefore, this embodiment of the invention proposes a data-driven approach to analyze the differences in power generation, the relationship between panel cleaning and power generation efficiency, and the differences in overall efficiency during the same period. This analysis determines the degree of impact of haze on each zone, and then modifies the current panel cleaning scheme based on this degree of impact. Figure 6 is an example flowchart of the modification of the panel cleaning scheme for the photovoltaic power plant. As shown in Figure 6, this process may include:

[0098] S2051. Analyze the difference between the measured daily power generation and the historical daily power generation of the photovoltaic power plant.

[0099] S2052. Analyze the correspondence between the panel cleaning degree data and power generation efficiency of each zone of the photovoltaic power plant.

[0100] In this embodiment of the invention, the photovoltaic power plant can collect the measured daily power generation and historical daily power generation within the plant. Additionally, it can obtain data on the cleaning degree of the panels in each zone and the corresponding power generation efficiency data. Through data statistics and processing, the difference between the measured daily power generation and the historical daily power generation of the photovoltaic power plant can be analyzed, and the correspondence between the panel cleaning degree data and the power generation efficiency in each zone of the photovoltaic power plant can be determined.

[0101] S2053. Based on the differences in power generation, analyze the obtained correspondence, the difference between the comprehensive efficiency of each zone of the photovoltaic power plant and the comprehensive efficiency of each zone in the same historical period, and the difference between the overall comprehensive efficiency of the photovoltaic power plant and the overall comprehensive efficiency in the same historical period, to obtain the degree of impact of haze on each zone.

[0102] To analyze the impact of haze on each zone, this step also analyzes the difference between the overall efficiency of each zone and its historical average for the same period, as well as the difference between the overall efficiency of the photovoltaic power plant and its historical average for the same period. Thus, the impact of haze on individual zones is analyzed from both the zone-level and overall perspectives of the photovoltaic power plant. For example, if a zone has low power generation efficiency and a low cleaning ratio, it is very likely that the low power generation efficiency is due to inadequate or untimely cleaning. The degree of impact of haze on a zone characterizes the extent to which haze interferes with the unsatisfactory power generation efficiency of that zone.

[0103] S2054. Modify the current board cleaning scheme based on the degree of impact of fog on each zone.

[0104] Based on the degree of impact of fog on each zone, component cleaning suggestion information is generated; then, the current component cleaning plan is modified based on the component cleaning suggestion information. The component cleaning suggestion information includes at least one of the following: area cleaning suggestion information, cleaning frequency suggestion information, cleaning speed suggestion information, and cleaning manpower allocation suggestion information.

[0105] The area cleaning recommendation information can provide suggestions on the cleaning ratio or cleaning priority for certain specific areas.

[0106] The cleaning frequency recommendation information can provide suggested cleaning frequencies, such as recommending cleaning once every two months, or increasing the cleaning frequency to once every two weeks during months with frequent smog.

[0107] The cleaning speed recommendation information can provide a suggested cleaning speed, such as recommending that all boards in the partition be cleaned within 10 days.

[0108] The cleaning manpower allocation suggestion can be a recommended manpower scale, such as increasing the number of cleaning personnel from 15 to 25 for each section, which would allow the power generation of that section to return to its ideal state more quickly.

[0109] S206. The photovoltaic power plant controls power generation based on the safety requirements of the power plant equipment and the received planned power generation curve.

[0110] If the second fitting judgment result is yes, then the actual power generation of the photovoltaic power plant can closely approximate the plant's own prediction. Since the first fitting judgment was performed, and the second fitting judgment only proceeds if the first result is yes, the prerequisite for executing S206 is that both fitting judgment results are yes. Under this premise, it means that the actual power generation can also match the grid's plan for the photovoltaic power plant. Next, the photovoltaic power plant can control its power generation based on the safety requirements of the plant's equipment and the received planned power generation curve, ensuring that the final power generation effect meets both the grid's planning and demand, while also maintaining the plant's operational needs and safety requirements.

[0111] Specifically, this step can begin with the following judgment:

[0112] Determine whether the photovoltaic power plant meets the safety requirements of its equipment if it performs active and reactive power conversion control during power generation according to the planned power generation curve. If the safety requirements are met, the photovoltaic power plant performs power generation control based on the planned power generation curve. If the safety requirements are not met, the photovoltaic power plant performs power generation control based on the planned power generation curve using peak shaving technology.

[0113] Specifically, the determination of safety requirements can be as follows: if the photovoltaic power plant performs active power and reactive power conversion control during the power generation process according to the planned power generation curve, whether the current and temperature of the photovoltaic power plant's equipment meet the safety requirements, and whether the operating voltage of the photovoltaic power plant's equipment is normal.

[0114] Current, temperature, and operating voltage are common parameters that can be collected when photovoltaic (PV) power plant equipment is operating. If a PV power plant is allowed to operate without considering the impact of fluctuations in these parameters on equipment safety, safety accidents are highly likely, and there is even a possibility of environmental damage. Furthermore, ensuring performance and safety directly impacts the power plant's power generation costs. If only power generation is considered without regard to equipment safety, frequent equipment failures may occur, leading to frequent equipment replacements and unnecessary equipment losses. Data regarding rated power, rated current, and rated operating temperature can be obtained from the equipment's factory specifications or user manual. In this embodiment of the invention, by analyzing the current, temperature, and operating voltage of the PV power plant equipment, it is determined whether the active and reactive power conversion control during power generation, according to the planned power generation curve, meets safety requirements. This ensures the stability and reliability of PV power generation in the power system.

[0115] In this embodiment of the invention, if the aforementioned safety requirements are not met, the photovoltaic power plant controls power generation based on the planned power generation curve using peak shaving technology. Generally, some photovoltaic power plants consider peak shaving by shutting down devices. However, according to the terms of take-or-pay and PPA (Power Purchase Agreement) contracts, a certain amount of reactive power needs to be generated. In low-voltage scenarios, reactive power generation directly and significantly impacts active power based on the inverter characteristic curve, thus affecting power generation and electricity costs. Therefore, by using artificial intelligence to digitally analyze the physical process, reactive power is adjusted to the period of maximum output capacity. Excess active power is converted into reactive power to meet contract requirements, minimizing reactive power generation during normal times and maximizing active power generation. Specifically, the conduction angle of the photovoltaic power plant's thyristors can be adjusted for peak shaving. Through this method, the current can be controlled, and the safety and lifespan of the equipment can be enhanced.

[0116] Before power generation control, the method further includes: analyzing the change characteristics of active power and reactive power of the inverter of the photovoltaic power plant under the influence of voltage factors, and determining the inflection point information in the change characteristic curves of active power and reactive power under different voltages;

[0117] Power generation control includes: controlling the reactive power of the inverter to be less than the value on the reactive power coordinate axis in the inflection point information. This controls the reactive power before reaching the inflection point, preventing active power from being significantly affected.

[0118] In this invention, by analyzing whether the predicted power generation curve fits the planned power generation curve, and in cases where a fit is not possible, the grid side continues to perform deep learning to adjust the planned power generation curve and resend it to the photovoltaic power plant. This enables the grid-side load plan to more stably align with the photovoltaic power plant's predicted power generation trend. By analyzing whether the measured power generation curve fits the predicted power generation curve and adjusting the photovoltaic power plant's panel cleaning scheme accordingly, the measured situation can be improved, allowing subsequent measured power generation curves to approximate the predicted power generation curve, greatly enhancing the determinism of photovoltaic power generation control in the power system. The photovoltaic power plant's power generation control is based on the safety requirements of the power plant equipment, ensuring equipment lifespan and performance. Therefore, this invention addresses the needs of both the power plant and the grid side, effectively overcoming the challenges of photovoltaic power generation control.

[0119] The following is a general description of the power generation control method based on active and reactive power adjustment of photovoltaic power generation provided by the technical solution of the present invention, with reference to Figure 7. Figure 7 involves the steps shown in Figure 2, and unlike Figure 2, Figure 7 also shows the process of generating the planned power generation curve and the predicted power generation curve on the grid side and the power plant side respectively before obtaining the planned power generation curve and the predicted power generation curve.

[0120] Furthermore, after completing the second fitting judgment (i.e., judging whether the measured power generation curve fits the predicted power generation curve), Figure 7 also shows the steps of the photovoltaic power plant sending power and cost data for the curtailment assistance function to the grid side, as well as the steps of the photovoltaic power plant receiving control commands from the grid side. These control commands are sent by the grid side after reaching a transaction agreement with the photovoltaic power plant based on the power and cost data. Therefore, if the photovoltaic power plant receives this control command, it means that both the photovoltaic power plant and the grid side have reached an agreement on the transaction based on the power and cost data. In this way, power generation control can be implemented in real time according to the specific requirements for active and reactive power in the planned power generation curve. During this control process, the safety requirements of the power plant equipment still need to be considered, and the reactive power generated needs to be controlled in conjunction with the changing characteristics of the inverter's active and reactive power.

[0121] In this embodiment of the invention, in conjunction with the above introduction to the power generation control of photovoltaic power plants, a process for overall power system control is also provided. Figure 8 shows the signaling diagram for overall power system control. This figure exemplarily illustrates the power plant-side components, including photovoltaic power plants, hydropower plants, and thermal power plants. The process for photovoltaic power plants has been specifically described above and will not be repeated here; it can be understood in conjunction with Figures 2 and 7. The power generation control processes for hydropower plants and thermal power plants are similar to those for photovoltaic power plants, both requiring comparison of their respective production factor data with corresponding historical data from the same period. For hydropower plants, specific constraints on some of their production factor data must also be considered. As shown in Figure 8, for photovoltaic power plants, hydropower plants, and thermal power plants, two fitting judgments are required to achieve stable and accurate control: a judgment on whether the predicted power generation of the power plant fits the planned power generation on the grid side, and a judgment on whether the measured power generation of the power plant fits the predicted power generation of the power plant. The generator sets operate according to the power generation control commands, and the data generated is incorporated into the historical database through digital analysis, serving as digital assets for subsequent power system scheduling and control.

[0122] The following section further introduces segmented control. Segmented control involves implementing virtual power plant functions for each photovoltaic (PV) power plant at different time periods through instructions issued by the power grid dispatch center (based on weather and digital twin data of the power grid). This provides relevant power factors for active and reactive power, adjusts the maximum power limits within each PV power plant, generates specific power factors and active and reactive power output values, and ultimately controls each PV power plant to convert excess active power into reactive power. The significance of segmented control lies in: 1. Maintaining power system stability by eliminating the need for PV power plants to generate reactive power via SVG (Static Var Generator), thus increasing equipment utilization; 2. Converting active power into reactive power, allowing excess active power to be utilized; 3. Strengthening the synergy between PV power plants and other power sources through virtual power plant aggregation and integration of resources; 4. Enhancing control flexibility, and the application of data twins allows for pre-defined segmented functions based on load, weather, and operating conditions, providing dispatch with an additional tool for managing active and reactive power within a given day; 5. Considering power grid and power plant safety, it effectively controls PV power plants as virtual power plants in the event of system failure or excessively low or high grid voltage, realizing the integration of artificial intelligence and the power grid. Segmented control is beneficial for peak shaving.

[0123] As an example, when performing segmented control, the power factor in the segmented function used from 5:30 to 9:30 is 0.95, and the power factor used in the segmented function from 9:30 to 14:00 is 0.97. Figure 9A shows a schematic diagram of the power factor setting for segmented control. The peak-shaving effect can be seen in Figure 9B; the left side of Figure 9B shows the peak-shaving effect before segmented control, and the right side shows the peak-shaving effect after segmented control.

[0124] Under the given power requirements, from 5:30 to 09:30 and from 14:00 to 19:00, a constant power factor of 0.95 is set to compensate for reactive power to achieve the required power factor value of 0.95.

[0125] To reduce active power loss, set the power factor to 0.97 from 09:30 to 14:00 PM, and carefully observe the inverter port voltage.

[0126] The following are some points to note regarding segmented control:

[0127] 1. Under the condition that the inverter's rated output voltage is ≥800Vac, the real-time reactive power should not exceed 26MVar. 2. If the inverter output voltage drops, pay attention to the change in the inverter port voltage value, and at the same time monitor the change in the reactive power value of the entire power station. Appropriately increase the power factor value to increase reactive power output. 3. The inverter's nighttime reactive power compensation function is disabled by default and needs to be set through management tools such as APP or data acquisition device before it can be used. Calculate the amount of reactive power that the power station needs to compensate for, and see how many inverters need to output reactive power at night. It is recommended to minimize the number of inverters that need to output reactive power at night. 4. When the nighttime reactive power compensation function needs to be enabled, the PID module must be used in the system, and the nighttime PID protection parameter setting must be enabled. 5. It is necessary to confirm with the main transformer manufacturer whether the voltage regulation mechanism can be frequently adjusted under conditions greater than 80% load.

[0128] Whether or not to increase active power output when output voltage drops depends on the cause of the voltage drop and the specific conditions of the power system. The following are some logic and principles:

[0129] 1. Relationship between active power and voltage: In a power system, the relationship between active power (P), voltage (U), and current (I) can be expressed by the formula P = UIcosφ (where cosφ is the power factor). If the load increases, the current I increases, and in order to keep the active power P constant, the voltage U must increase accordingly. If the voltage U decreases, then in order to maintain the same active power output, the current I must increase, which may lead to increased line losses and a further drop in voltage.

[0130] 2. Causes of voltage dips: Voltage dips can be caused by a variety of factors, including a sudden increase in grid load, a decrease in generator output, and transmission line faults. If the voltage dip is due to an increase in load, it may be necessary to increase the active power output of the generator to maintain grid voltage stability.

[0131] 3. Impact of Reactive Power: Reactive power has a more direct impact on voltage. Insufficient reactive power leads to voltage drop because inductors consume reactive power, causing direct-axis demagnetizing armature reactions in synchronous generators, reducing magnetic flux, and thus lowering induced electromotive force and voltage. Therefore, increasing reactive power output can help maintain or restore voltage levels.

[0132] 4. Power Grid Stability: The stability of a power grid depends not only on active power but also on reactive power. During voltage dips, adjustments to both active and reactive power may need to be considered simultaneously to ensure stable grid operation.

[0133] 5. Dynamic Voltage Restorer (DVR): In some cases, devices such as DVRs can provide temporary voltage support when voltage drops occur, reducing reliance on active power output.

[0134] In summary, whether or not to increase active power output when output voltage drops requires analysis based on the specific circumstances. Generally, handling voltage drops requires comprehensive consideration of adjustments to both active and reactive power, as well as the stability and security of the power grid. In practice, grid operators will determine how to adjust power generation and grid operating parameters to cope with voltage drops based on real-time data and predictive models.

[0135] Based on the methods described in the foregoing embodiments, the present invention also provides a power generation control device based on active and reactive power adjustment of photovoltaic power generation. This will be described below with reference to Figure 10.

[0136] Figure 10 is a schematic diagram of the power generation control device based on active and reactive power adjustment of photovoltaic power generation according to an embodiment of the present invention. As shown in Figure 10, the device includes:

[0137] The curve acquisition module is used to acquire the planned power generation curves of active and reactive power allocated to the photovoltaic power plant by the grid side, and to acquire the predicted power generation curves of active and reactive power generated by the photovoltaic power plant itself.

[0138] The fitting judgment module is used to determine whether the predicted power generation curve fits the planned power generation curve.

[0139] The curve modification module is used to modify the planned power generation curve through deep learning if the predicted power generation curve cannot be fitted to the planned power generation curve, and then resend the modified curve to the photovoltaic power plant.

[0140] The fitting judgment module is further used to determine whether the measured active power and reactive power curves of the photovoltaic power plant fit the predicted power generation curve if the predicted power generation curve fits the planned power generation curve.

[0141] The cleaning scheme modification module is used to modify the panel cleaning scheme of the photovoltaic power plant if the measured power generation curve and the predicted power generation curve cannot be fitted.

[0142] The power generation control module is used to control the power generation of the photovoltaic power plant based on the safety requirements of the power plant equipment and the received planned power generation curve if the measured power generation curve fits the predicted power generation curve.

[0143] In the optional implementation, the power generation control module is specifically used for:

[0144] Determine whether the photovoltaic power plant meets the safety requirements of its equipment if it performs active and reactive power conversion control during the power generation process according to the planned power generation curve.

[0145] If the safety requirements are met, the photovoltaic power plant controls power generation based on the planned power generation curve; if the safety requirements are not met, the photovoltaic power plant controls power generation based on the planned power generation curve using peak shaving technology.

[0146] In the optional implementation, the power generation control module is specifically used for:

[0147] If the photovoltaic power plant performs active and reactive power conversion control during the power generation process according to the planned power generation curve, determine whether the current and temperature of the photovoltaic power plant's equipment meet safety requirements, and whether the operating voltage of the photovoltaic power plant's equipment is normal.

[0148] In the optional implementation, the power generation control module is specifically used for:

[0149] Adjust the conduction angle of the thyristor in the photovoltaic power plant to perform peak reduction.

[0150] In an optional implementation, the power generation control device based on photovoltaic power generation active and reactive power adjustment also includes an inflection point information determination module, which is used to analyze the change characteristics of active power and reactive power of the inverter of the photovoltaic power plant under the influence of voltage factors before power generation control, and determine the inflection point information in the change characteristic curves of active power and reactive power under different voltages. The inflection point information includes the value on the reactive power coordinate axis and the value on the active power coordinate axis.

[0151] The power generation control module is specifically used for:

[0152] The reactive power value of the inverter is controlled to be less than the reactive power coordinate axis value in the inflection point information.

[0153] In the optional implementation, the curve modification module is specifically used for:

[0154] The grid-side deep learning method learns the difference between the measured daily power generation of the photovoltaic power plant and the historical daily power generation, and learns the difference between the comprehensive efficiency of the photovoltaic power plant and the comprehensive efficiency of the same period in history. It also combines the panel cleaning degree data of the photovoltaic power plant to identify the impact of haze and adjust the planned power generation curve.

[0155] In the optional implementation, the cleaning scheme modification module is specifically used for:

[0156] Analyze the difference between the measured daily power generation and the historical daily power generation of the photovoltaic power plant;

[0157] Analyze the correlation between the panel cleaning level data and power generation efficiency in each zone of the photovoltaic power plant;

[0158] Based on the differences in power generation, the corresponding relationship obtained by analysis is obtained, the difference between the comprehensive efficiency of each zone of the photovoltaic power plant and the comprehensive efficiency of each zone in the same period of history, and the difference between the overall comprehensive efficiency of the photovoltaic power plant and the overall comprehensive efficiency in the same period of history, so as to obtain the degree of impact of haze on each zone.

[0159] The current board cleaning plan has been modified based on the degree of impact of fog on each zone.

[0160] In the optional implementation, the cleaning scheme modification module is specifically used for:

[0161] Based on the degree of impact of fog on each zone, panel cleaning recommendations are generated; the panel cleaning recommendations include at least one of the following: area cleaning recommendations, cleaning frequency recommendations, cleaning speed recommendations, and cleaning manpower allocation recommendations;

[0162] Modify the current board cleaning plan based on the board cleaning suggestion information.

[0163] In an optional implementation, the power generation control device based on photovoltaic power generation active and reactive power adjustment further includes a first curve generation module, used to generate planned power generation curves for active and reactive power allocated to the photovoltaic power plant on the grid side. The first curve generation module is specifically used for:

[0164] Based on the characteristic production factor data of each power plant and the control production factor data of the power grid, the grid side determines the load curve to be allocated to each power plant, including the planned power generation curves of active power and reactive power allocated to photovoltaic power plants. The active power control frequency is used, and the reactive power control voltage is used.

[0165] In an optional implementation, the power generation control device based on photovoltaic power generation active and reactive power adjustment further includes a second curve generation module, used by the photovoltaic power plant to generate predicted power generation curves for active and reactive power. The second curve generation module is specifically used for:

[0166] The photovoltaic power plant generates predicted power generation curves for active and reactive power based on equipment performance, weather data, and historical power generation for the same period. The equipment performance includes the characteristics of the inverter's active and reactive power changes under the influence of voltage factors.

[0167] In an optional implementation, the power generation control device based on the adjustment of active and reactive power of photovoltaic power generation further includes a transmitting module, which is used to transmit the power generation data and cost data of the curtailment auxiliary function to the grid side before the photovoltaic power plant performs power generation control based on the safety requirements of the power plant equipment and the received planned power generation curve.

[0168] The power generation control module is specifically used to control power generation based on the safety requirements of the power plant equipment and the received planned power generation curve if the photovoltaic power plant receives a control command from the grid side; the control command is sent by the grid side after reaching a transaction agreement with the photovoltaic power plant based on the power data and the cost data.

[0169] It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments. The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separate. The components indicated as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment solution according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0170] The above description is merely one specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A power generation control method based on photovoltaic power generation active and reactive power adjustment, characterized in that, include: Obtain the planned power generation curves of active and reactive power allocated to the photovoltaic power plant by the grid, and obtain the predicted power generation curves of active and reactive power generated by the photovoltaic power plant itself. Determine whether the predicted power generation curve fits the planned power generation curve; If the predicted power generation curve cannot be fitted to the planned power generation curve, the grid side modifies the planned power generation curve through deep learning and resends it to the photovoltaic power plant after the modification is completed; if the predicted power generation curve fits the planned power generation curve, it is determined whether the measured power generation curves of the photovoltaic power plant's active power and reactive power fit the predicted power generation curve. If the measured power generation curve cannot be matched with the predicted power generation curve, the panel cleaning scheme of the photovoltaic power plant is modified; if the measured power generation curve matches the predicted power generation curve, the photovoltaic power plant controls power generation based on the safety requirements of the power plant equipment and the received planned power generation curve.

2. The method according to claim 1, characterized in that, The photovoltaic power plant controls power generation based on the safety requirements of the power plant equipment and the received planned power generation curve, including: Determine whether the photovoltaic power plant meets the safety requirements of its equipment if it performs active and reactive power conversion control during the power generation process according to the planned power generation curve. If the safety requirements are met, the photovoltaic power plant controls power generation based on the planned power generation curve; if the safety requirements are not met, the photovoltaic power plant controls power generation based on the planned power generation curve using peak shaving technology.

3. The method according to claim 2, characterized in that, The determination of whether the photovoltaic power plant meets the safety requirements of its equipment if it performs active and reactive power conversion control during power generation according to the planned power generation curve includes: If the photovoltaic power plant performs active and reactive power conversion control during the power generation process according to the planned power generation curve, determine whether the current and temperature of the photovoltaic power plant's equipment meet safety requirements, and whether the operating voltage of the photovoltaic power plant's equipment is normal.

4. The method according to claim 2, characterized in that, The photovoltaic power plant controls power generation based on the planned power generation curve using peak shaving technology, including: Adjust the conduction angle of the thyristor in the photovoltaic power plant to perform peak reduction.

5. The method according to claim 1, characterized in that, Before performing power generation control, the method further includes: The characteristics of active and reactive power variation of the inverter of the photovoltaic power plant under the influence of voltage factors are analyzed, and the inflection point information in the curves of active and reactive power variation characteristics under different voltages is determined. The inflection point information includes the values ​​on the reactive power axis and the active power axis. Power generation control includes: The reactive power value of the inverter is controlled to be less than the reactive power coordinate axis value in the inflection point information.

6. The method according to claim 1, characterized in that, The grid side modifies the planned power generation curve through deep learning, including: The grid-side deep learning method learns the difference between the measured daily power generation of the photovoltaic power plant and the historical daily power generation, and learns the difference between the comprehensive efficiency of the photovoltaic power plant and the comprehensive efficiency of the same period in history. It also combines the panel cleaning degree data of the photovoltaic power plant to identify the impact of haze and adjust the planned power generation curve.

7. The method according to claim 1, characterized in that, The modification of the photovoltaic power plant's panel cleaning solution includes: Analyze the difference between the measured daily power generation and the historical daily power generation of the photovoltaic power plant; Analyze the correlation between the panel cleaning level data and power generation efficiency in each zone of the photovoltaic power plant; Based on the differences in power generation, the corresponding relationship obtained by analysis is obtained, the difference between the comprehensive efficiency of each zone of the photovoltaic power plant and the comprehensive efficiency of each zone in the same period of history, and the difference between the overall comprehensive efficiency of the photovoltaic power plant and the overall comprehensive efficiency in the same period of history, so as to obtain the degree of impact of haze on each zone. The current board cleaning plan has been modified based on the degree of impact of fog on each zone.

8. The method according to claim 7, characterized in that, The modification of the current board cleaning scheme based on the degree of impact of fog on each zone includes: Based on the degree of impact of fog on each zone, panel cleaning recommendations are generated; the panel cleaning recommendations include at least one of the following: area cleaning recommendations, cleaning frequency recommendations, cleaning speed recommendations, and cleaning manpower allocation recommendations; Modify the current board cleaning plan based on the board cleaning suggestion information.

9. The method according to claim 1, characterized in that, The planned power generation curves for active and reactive power allocated to the photovoltaic power plant by the grid side are generated in the following way: Based on the characteristic production factor data of each power plant and the control production factor data of the power grid, the grid side determines the load curve to be allocated to each power plant, including the planned power generation curves of active power and reactive power allocated to photovoltaic power plants. The active power control frequency is used, and the reactive power control voltage is used.

10. The method according to claim 1, characterized in that, The methods for generating predicted power generation curves for active and reactive power by the photovoltaic power plant include: The photovoltaic power plant generates predicted power generation curves for active and reactive power based on equipment performance, weather data, and historical power generation for the same period. The equipment performance includes the characteristics of the inverter's active and reactive power changes under the influence of voltage factors.

11. The method according to claim 1, characterized in that, Before the photovoltaic power plant performs power generation control based on the safety requirements of the power plant equipment and the received planned power generation curve, the method further includes: The photovoltaic power plant sends power and cost data for the curtailment assistance function to the grid side. If the photovoltaic power plant receives a control command from the grid side, it will control power generation based on the safety requirements of the power plant equipment and the received planned power generation curve; the control command is sent by the grid side after reaching a transaction agreement with the photovoltaic power plant based on the power data and the cost data.

12. A power generation control device based on active and reactive power adjustment of photovoltaic power generation, characterized in that, include: The curve acquisition module is used to acquire the planned power generation curves of active and reactive power allocated to the photovoltaic power plant by the grid side, and to acquire the predicted power generation curves of active and reactive power generated by the photovoltaic power plant itself. The fitting judgment module is used to determine whether the predicted power generation curve fits the planned power generation curve. The curve modification module is used to modify the planned power generation curve through deep learning if the predicted power generation curve cannot be fitted to the planned power generation curve, and then resend the modified curve to the photovoltaic power plant. The fitting judgment module is further used to determine whether the measured active power and reactive power curves of the photovoltaic power plant fit the predicted power generation curve if the predicted power generation curve fits the planned power generation curve. The cleaning scheme modification module is used to modify the panel cleaning scheme of the photovoltaic power plant if the measured power generation curve and the predicted power generation curve cannot be fitted. The power generation control module is used to control the power generation of the photovoltaic power plant based on the safety requirements of the power plant equipment and the received planned power generation curve if the measured power generation curve fits the predicted power generation curve.

Images